IN VIVO EXAMINATION OF THE MOLECULAR MECHANICS UNDERLYING APICAL

CONSTRICTION’S INITIATION IN C. ELEGANS GASTRULATION

Timothy Dennison Cupp

A thesis submitted to the faculty of the University of North Carolina at Chapel Hill in partial fulfillments of the requirements for the degree of a Master in Science in the

Department of Cell Biology and Physiology in the School of Medicine.

Chapel Hill

2016

Approved by:

Bob Goldstein

Keith Burridge

Richard Cheney

Stephanie Gupton

Amy Maddox

brought to you by COREView metadata, citation and similar papers at core.ac.uk

provided by Carolina Digital Repository

ii

© 2016 Timothy Dennison Cupp

ALL RIGHTS RESERVED

iii

ABSTRACT

Timothy Dennison Cupp: In vivo examination of the molecular mechanics underlying apical constriction’s initiation in C. elegans gastrulation

(Under the direction of Bob Goldstein)

One remarkable finding from research in developmental biology is that

surprisingly few cellular behaviors are responsible for the wide variety of morphogenetic

events common among all eukaryotes. Molecular mechanisms underlying cell shape

changes during tissue restructuring can explain how morphogenesis proceeds in vivo.

During apical constriction, contractile myosin movements become linked to apical

junctions, resulting in junctional pulling that can change cell shape. The process by

which this dynamic linkage is achieved remains unknown, though it occurs with strict

developmental timing in at least two systems. Since timing and patterning information

instruct enrichment of specific mRNAs in the cells that apically constrict in C. elegans,

we targeted these genes in an RNAi screen, identifying candidates that have an

involvement in apical constriction. Our data suggest that zyxin, an important Focal

Adhesion protein, may mediate the connections between the actomyosin cortex and

adherens junctions during the initiation of apical constriction.

iv

TABLE OF CONTENTS

LIST OF TABLES..............………......………………………………………………………..vi

LIST OF FIGURES……………………..……………………………….......................……..vii

LIST OF ABBREVIATIONS............................................................................................viii

CHAPTER 1: IN VIVO EXAMINATION OF APICAL CONSTRICTION............................1

INTRODUCTION……….....…………………………………………………………....1

EXPERIMENTAL APPROACH and METHODS ………..…………………………..6

RESULTS..............................................................................................................8

DISCUSSION…………………………………………………………………………..13

Future Directions...…………………………………………………………….15

APPENDIX 1: TABLE 1 – RESULTS OF RNAi SCREEN…………………………….......19

APPENDIX 2: FIGURES…………….…………………………………………………..……22

Figure 1…………………………………………………………………………………22

Figure 2…………………………………………………………………………………23

Figure 3…………………………………………………………………………………24

Figure 4…………………………………………………………………………………25

Figure 5…………………………………………………………………………………26

Figure 6………………………………………………………………………………....27

Figure 7…………………………………………………………………………………28

Figure 8…………………………………………………………………………………29

v

Figure 9…………………………………………………………………………………30

REFERENCES…………………………………………………………………………………31

vi

LIST OF TABLES

Table 1 – Results of RNAi Screen…………………………………………………………..21

vii

LIST OF FIGURES

Figure 1 – Apical constriction is a conserved process in C. elegans gastrulation as well as vertebrate neural tube formation……………………………………………….…………22

Figure 2 – C. elegans embryos achieve gastrulation via

apical constriction…………………………………………………………..….23

Figure 3 – Pseudo-kymograph of apical myosin (green) and membrane (red) dynamics during apical constriction……………………...24

Figure 4 – The temporally-regulated link between adherens

junctions and F-actin…………………………………………………………..25

Figure 5 – Choosing candidates based on expression profile in the early embryo…………………………………………………………….…26

Figure 6 – Results of RNAi Screen………………………………………………………….27

Figure 7 – E cells divide at ventral surface in Gad embryos……………………………..28

Figure 8 – Slippage Rate remains abnormally high after zyxin depletion………………………………………………………………...….......29

Figure 9 – Our model of zyxin’s action during apical constriction……….…..…….….....30

viii

LIST OF ABBREVIATIONS

C. elegans Caenorhabditis elegans

CCC Cadherin-Catenin Complex

dsRNAs Double-stranded ribonucleic acids

E cell Endodermal Precursor Cell

Focal Adhesions FAs

Focal Adherens Junctions FAJs

GFP Green Fluorescent Protein

Gad Gastrulation defective

MS cell Mesodermal Precurosor Cell

TagRFP Tag Red Fluorescent Protien

RacGEF Guanine Exchange Factor for Rac (GTPase)

RNAi Ribonucleic Acid Interference

RPKM Reads Per Kilobase per Millio

1

CHAPTER 1: IN VIVO EXAMINATION OF APICAL CONSTRICTION. INTRODUCTION

Apical constriction is a cell shape change that can drive tissue morphogenesis in

nearly all metazoans3. Vertebrates utilize apical constriction to direct neural tube closure

(Figure 1) and roughly 300,000 newborns suffer from neural tube closure defects

worldwide annually14. During this developmentally regulated event, the tension arising

from contractions within the actomyosin cortex is transmitted across cell junctions to pull

on neighboring cells. The resulting change in cell shape ultimately drives tissue folding

and invagination3,4. Identifying the key molecules involved and discerning how they

behave during apical constriction is crucial to our understanding of this morphogenetic

event. Our lab has identified a number of genes involved in apical constriction during C.

elegans gastrulation, but the precise details of the mechanism are still murky8,17,21. The

adhesive cadherin-catenin complex (CCC), containing alpha-catenin, beta-catenin, and

E-cadherin, becomes apically enriched in E cells upon myosin activation and is required

for apical constriction to proceed, somehow dynamically connecting to the tensile

cytoskeleton with proper timing. The exact protein-protein interactions that facilitate

force transduction between the cytoskeleton and the CCC has been under intense

debate in recent years (Figure 4)30,31. Some camps hold that actin can directly bind to

alpha-catenin under certain conditions, though these experiments typically carry

caveats stemming from their in vitro methodologies32. Here, we consider potential

interactions that enable protein binding in a developing organism.

2

In C. elegans, two endodermal precursor cells (E cells) originate at the surface of

the embryo when the E cell progenitor divides. Gastrulation begins at the 26- to 28-cell

stage as the E cells apically constrict and internalize (Figure 1, Figure 2B). C. elegans

is well-suited for identifying and understanding the complex roles of candidate molecular

triggers for apical constriction in vivo. Several factors make C. elegans an attractive

candidate for this type of in vivo morphogenesis research. The organism’s genetic

tractability allows for direct edits to the genome, making direct, specific mutagenesis

and fluorescent protein fusions a relatively straightforward task15,16. C. elegans embryos

are also transparent, permitting direct observation of fluorescently-tagged protein

dynamics throughout development. Because of the animal’s short generation time, we

can identify, tag, and image any protein of interest all within the span of a month. Since

apical constriction progresses in such a spatiotemporally-stereotyped manner and

because our model provides many in vivo experimental advantages, C. elegans is

valuable for studying morphogenesis.

Developmental patterning is integrated with spatial information in E cells,

instructing their decision to undergo apical constriction at the 26-28 cell stage in the

early C. elegans embryo with precise timing. Expression of the end-3 transcription factor

confers an endodermal fate to these cells, which promotes the production and apical

recruitment of the myosin light-chain kinase, MRCK-1, to apical contact-free cell

surfaces17.

MRCK-1’s apical recruitment in E cells results in phosphorylation of myosin’s

regulatory light chain and its subsequent activation at the apical surface. The apical

enrichment and activation of myosin provides the cortical tension that constriction

3

requires to proceed. Previous hypotheses posited that this sudden increase in tension

within cells would alone be sufficient to initiate constriction. While the activation and

apical localization of myosin is absolutely required for constriction to proceed11,12, our

lab has shown that these events are not themselves the immediate trigger of apical

constriction in either Drosophila or C. elegans5. In fact, myosin’s apical enrichment and

activation precede the initiation of constriction by several minutes. Myosin activity

seems to stabilize the adhesive structures at apical surfaces of cell-cell junctions by

preventing the sequestration/internalization of cadherin-catenin complexes, but does not

directly initiate tissue rearrangement19,26.

During this period of increased tension within E cells’ apical surfaces before the

onset of apical constriction5,17 myosin puncta flow centripetally along the cell’s apical

cortex without accompanying movement of associated cell membranes (Figure 3). We

refer to this phenomenon as “slippage.” With precise timing, these myosin movements

“couple” with the cell-cell junctions and apical constriction proceeds, suggesting that

tension is transmitted through cell junctions to cytoskeleta of neighboring cells. During

coupling, slippage is almost entirely eliminated. The precise mechanism of how myosin

and membrane structures achieve this coupling with temporal accuracy remains a

mystery.

One hypothesis that can explain this event’s nature is the missing link model

(Figure 4). In the missing link model, there is a disconnect between the actomyosin

cortex and the cadherin-catenin complex of apical adherens junctions. According to this

hypothesis, to establish a connection and accomplish apical constriction, some crucial

protein (or set of proteins) is expressed, recruited, post-translationally modified, or

4

otherwise made available to interface between the actomyosin cortex and adherens

junctions. The formation of this linkage results in force transmission between

neighboring cells and constriction can occur.

To test this hypothesis and determine which proteins might be involved in

forming this developmentally-regulated linkage, we decided to perform an RNAi screen.

A single-cell transcriptomic analysis of the early C. elegans embryo provided the basis

of our candidate list9. In this analysis, expression levels of each transcript were

measured in each individual cell through the 16-cell stage of development. With this

information, we constructed a candidate list of genes most highly enriched in E cells (16

cell stage) and their progenitor (8 cell stage). We hypothesized that at least one of the

candidates in this list will be involved in forming a connection between adherens

junctions and the contractile actomyosin network.

To determine which proteins from our candidate list might be involved in apical

constriction, we designed the screen to seek defects in C. elegans gastrulation

(Gastrulation defective = Gad phenotype). We injected dsRNA constructs targeting the

candidate gene’s mRNA into mature adult worm gonads. Injection of dsRNA leads to

highly penetrant knockdown effects (as compared to RNAi feeding strategies)27. After

knockdown, we established that an embryo is Gad if the E cells fail to fully internalize

into the blastocoel before completing their first division. From our candidate list, we

isolated a number of genetic targets that result in a Gad phenotype after knockdown in

wild-type embryos. In particular, targeting zyxin (zyx-1) transcripts with dsRNAs resulted

in a high incidence of Gad embryos in both wild-type and sensitized backgrounds. Due

5

to this phenotype and zyxin’s mRNA enrichment in E cells around the time of

gastrulation, we sought to consider zyxin’s role in apical constriction.

By filming embryos expressing fluorescently-tagged proteins, we were able to

track the movements of myosin puncta in relation to cellular borders during apical

constriction. We depleted these embryos of zyxin mRNAs to scrutinize these molecular

dynamics. In constricting cells, myosin puncta flow centripetally along the apical

surface, apparently supplying tension to the cortical network3,19,20. As shown in Figure

3, wild-type E cells experience several minutes of slippage during which myosin flows

without accompanying movement of the cell junctions. These cells then initiate

constriction with precise timing and coupling occurs. After targeting zyxin for

knockdown, we found that the slippage rate erroneously remained high during the latter

stages of gastrulation. We speculate that this sustained slippage explains the Gad

phenotype in zyxin-depleted embryos.

Zyxin has long been appreciated for its role in focal adhesion maturation in cells

crawling along rigid surfaces, acting as a mechano-sensitive adapter protein between

the integrin-signaling and cytoskeletal layers10-13,18 We show here that zyxin is important

even in the absence of a rigid substrate. Our data suggest that zyxin may have an

additional role in strengthening the connections between the cytoskeleton and adherens

junctions. Future experiments revealing zyxin’s expression level, timing, recruitment,

and localization will be instrumental in confirming its function during apical constriction.

6

EXPERIMENTAL APPROACH and METHODS RNA interference

We injected 1 µg/µL of dsRNA in TE Buffer into the gonads of young adult

worms. Each dsRNA construct was designed to target the first kilobase of each

candidate gene’s exonic code. After 36 hours incubation at 20°C, we dissected out

dsRNA-treated embryos and mounted them laterally on glass coverslips no later than

the 8-cell stage.

DIC and fluorescence microscopy

We filmed each embryo under DIC illumination at one minute intervals for 1.5

hours to measure its progression through gastrulation. In typical wild-type embryos, E

cells don’t divide until they have internalized entirely into the blastocoel and are fully

covered by their neighbors. We consider an embryo to be Gad if this division occurs

before the completion of E cell internalization.

Candidates with an apparent effect on gastrulation were further tested to

establish any potential roles in apical constriction. DIC illumination of laterally mounted

embryos can only tell us whether gastrulation has been successful (Figure 6), but does

not directly reveal the movements of apical components. To reveal these movements,

we mounted embryos on their ventral surface so that the apical surfaces of E cells are

exposed to the glass coverslip and imaged the ventral surface. Spinning disk confocal

microscopy allowed collection of images of the entire apical surface of the E cell (Figure

7A,B). Utilizing worms expressing a myosin-GFP fusion plus a PH domain-mCherry

fusion (see Strains and Worm Maintenance, below), we were able to track their

movements of myosin and membrane (into which the PH domain embeds itself) with

7

high temporal resolution (1 image every 3 seconds for 5 minutes), in order to measure

the slippage rate during early and late stages of gastrulation. The MTrackJ plugin in

ImageJ was used to track myosin and membrane movements and calculate their

velocities.

Strains and worm maintenance

Nematodes with cultured and handled as described28. The following reporter and

mutant strains were used: MT4417 ced-5(n1812); SU348 sax-7(eq1); LP54

mCherry::PH; NMY-2::GFP. Imaging was performed at 20°C – 23°C for all strains listed.

8

RESULTS Determining Candidate List for RNAi Screen

From the dozens of enriched transcripts in the E cells and the E cell progenitor,

we considered three classes of genes: (1) genes whose predicted products are well-

characterized and might serve a mechanical purpose; (2) genes with only a few

predicted domains; and (3) genes whose products are likely not involved in apical

constriction at all. We predict that products in the first class will either interact with the

cytoskeleton directly or otherwise sense and transduce force via stretch-induced binding

(perhaps revealing cryptic domains as in vinculin, talin) and signaling.

The second class of products only have a few domains listed, merely hinting at

prospective functions. Their conceivable involvements in apical constriction range from

cell-cell adhesion (e.g. C-type lectin fold protein encoded by F25D7.2) to cell signaling

(e.g. tyrosine kinase and phosphatase domains are common features among

candidates). Though these proteins aren’t well-characterized, we can still offer simple

hypotheses positing roles in gastrulation.

The final class of genes encoded lysosomal proteins, glycosylation proteins,

Argonaut proteins, etc. and will likely have little direct involvement in the progression of

apical constriction since they do not have any proposed mechanical function. We

decided to exclude this set of genes was excluded from the screen. The final non-

comprehensive list includes 28 candidate genes.

After being targeted for depletion, several candidates display a Gad phenotype

After treating embryos with dsRNAs targeting the candidate genes, we assayed

embryos for their ability to progress through gastrulation3,4. E cells and the neighboring

9

MS cells (mesodermal precursor cells) are born within one minute of each other. These

MS cells divide into four daughters 25 minutes following their birth, followed shortly

thereafter by E cell apical constriction and internalization. In untreated wild-type

embryos, E cells fully internalized in 16.2 (+/- 2.0) minutes following MS cell divisions

(Table 1, Figure 6). After completing internalization (when no part of either E cell is

exposed to the embryo’s ventral surface), each E cell divided in 3.7 (+/- 1.7) minutes.

We found that targeting certain genes for knockdown often led to delays in

internalization. In some cases, E cells failed to completely internalize before undergoing

division (Figure 7). This is nearly always due to a delay in internalization rather than a

defect in cell division timing (Table 1, Figure 6).

From the candidates tested, we isolated several genes which have an effect on

gastrulation in the wild-type background. In addition to severely delaying E cell

internalization, treating early embryos with dsRNAs also led to a > 25% incidence of the

Gad phenotype for 9/23 candidates tested in the wild-type background (See Table 1,

Figure 6). To overcome potential genetic redundancies, we also targeted some

candidates for knockdown in sensitized genetic backgrounds (See Experimental

Approaches and Methods). Ced-5, which encodes a RacGEF38, acts redundantly with

hmr-1 (encoding cadherin) during C. elegans gastrulation21. Treatment of dsRNAs in the

ced-5 genetic null background led to a very high incidence of Gad phenotypes; 4/6

candidates were Gad in more than half of the embryos. SAX-7 is a transmembrane cell

adhesion receptor molecule29 that functions partially redundantly with cadherin during

gastrulation37. 8 of the 13 candidates targeted for depletion in the sax-7 null background

displayed Gad phenotype in > 25% of embryos. We speculate that these candidates

10

function in a pathway with cadherin to allow its physical linkage to cytoskeletal

elements.

In comparison to earlier screening methods21, our screen yielded in a high

number of promising candidates with highly penetrant effects on gastrulation. Since

these genes have an apparent effect on gastrulation, we wanted to uncover any direct

roles in apical constriction. One promising candidate has been chosen for further

analysis so far.

Determining zyxin’s role in promoting completion of gastrulation

Knockdown of zyxin resulted in a noticeable phenotype in both wild-type and

sensitized backgrounds. It has also been previously shown to have a role in mechano-

sensitive actions. Zyxin is mainly appreciated for its role in biomechanical feedback

during focal adhesion maturation10,12, though recent evidence suggests it can also work

at cell-cell contacts24.

Zyxin is an attractive candidate for additional study for a number of reasons. For

one, zyxin’s expression pattern is among the most striking of all the candidates in terms

of its E cell enrichment. These transcripts are 358 times more enriched in the E cell

progenitor than in its neighbors (Table 1, Figure 5B). Additionally, the effect of zyxin

targeting on the early embryo is among the strongest that we observed for well-

characterized proteins (Table 1). Zyxin’s known biomechanical function also offered a

clear set of hypotheses detailing how it might function to initiate apical constriction. For

these reasons, we predict that zyxin contributes to a developmentally-regulated

assemblage that links actomyosin and adherens junctions. Loss of zyxin would weaken

11

this connection, thereby preventing membranes from moving in tangent with myosin

puncta.

Our gastrulation data raise the possibility that zyxin operates in apical

constriction, but the screening method does not allow direct detection of this

phenomenon. In our screen, we filmed laterally mounted embryos under DIC

illumination. The apical surfaces of E cells face the embryo’s ventral surface, so lateral

mounts, while quick and technically easy to perform, are good for screens but do not

allow for direct observation of constricting apical membranes.

High-speed fluorescent movies of ventrally mounted embryos allow us to

measure myosin’s rate of flow as well the adjacent cell membrane’s rate of centripetal

movement simultaneously. We define the difference in speeds as “slippage” between

myosin and membrane. As in previous analyses, we define two stages of apical

constriction for convenience: The early stage, a period spanning the 10 minutes

following MS daughter division; and the subsequent late stage, when E cell apical

surfaces actually constrict5. During the early stage, apical myosin flows centripetally

without accompanying movement of associated membranes. In short, there is a high

degree of slippage. As constriction begins, the membrane slowly begins to move at a

similar rate as myosin. During this late stage, the slippage rate gradually falls to under

0.5 µm/min in control embryos.

After treating embryos with dsRNAs targeting zyxin, we found that there was no

change in myosin’s rate of centripetal apical flow in E cells during either the early stage

or late stages. During the late stage, however, myosin movements fail to completely

couple with membrane movements in each case. As a result, the average slippage rate

12

remained significantly high (Figure 8). We conclude that depletion of zyxin in the early

C. elegans embryo prevents the tensile actomyosin cortex from linking to adhesive

junctional components, leading to defects in apical constriction and thus in gastrulation.

Zyxin’s expression pattern and timing suggest that it may be temporally regulated by

expression timing, and together with these data, hints at a developmentally regulated

role in tissue morphogenesis.

13

DISCUSSION

C. elegans embryos employ a basic cellular behavior to drive gastrulation. This

behavior, apical constriction, drives tissue rearrangement in many different biological

systems4,5. During apical constriction, myosin contractions elicit tension within the

actomyosin cortex. In a constricting cell, this tensile network somehow becomes

mechanically connected to cellular junctions, and neighboring cells are pulled over its

apical surface, inducing its internalization. The progression of this event is highly

stereotypical and appears to be under tight developmental regulation in worms4,6,17.

However, we can disrupt this process if we deplete early embryos of certain genetic

factors. One such gene which appears to be vital to the progression of apical

constriction is zyxin.

Direct injection of dsRNAs into adult gonads (in favor of feeding, where

knockdown is less effective) and the unique construction of the RNAi candidate list

yielded more penetrant effects on gastrulation than previous screening efforts21. After

targeting zyxin for knockdown with dsRNAs for 36 hours, many embryos demonstrated

a Gad phenotype. These defects arose in both sensitized (sax-7 null, ced-5 null) and

wild-type backgrounds. Knockdown of other cytoskeleton-related transcripts, including

alpha-Catulin (ctn-1), girdin (grdn-1), and formin (cyk-1), also led to an increased

incidence of Gad embryos (Table 1, Figure 6). Subsequent high temporal resolution

imaging has allowed us to directly observe the molecular dynamics at play within

constricting cells. These movies reveal that the Gad phenotype detected in zyxin-

knocked down embryos is likely due to a defect in membrane-cytoskeleton coupling, as

hypothesized. In these cases, slippage rates between myosin and membrane

14

movements remain high abnormally late in the process. That is, myosin contractile

movements do not efficiently couple with the cellular junctions during the late stages of

gastrulation as they should. We hypothesize that zyxin’s involvement in apical

constriction is at the interface between the contractile cytoskeleton and components

within adherens junctions at the apical surface.

The classic literature on zyxin presents this protein as a mechano-sensitive

element that is chiefly involved in tension sensing at integrin-based focal adhesions

during cellular migration10,12. Its stretch-induced recruitment to focal adhesions is

required for the subsequent recruitment of other cytoskeleton-associated proteins, such

as Ena/VASP and testin11. The ultimate consequence of stretch and thus zyxin

recruitment appears to be the strengthening actin cables’ attachment to adhesive

structures on a rigid substrate. Recent evidence suggests that zyxin not only has a role

in focal adhesions, but can also be found at the interface between two cells24.

In a set of in vitro experiments24, Oldenberg et al. pharmacologically induced

stretch and used super resolution microscopy to view several members of the focal

adhesion complex. They found that stretch induction results in the formation of “Focal

Adherens Junctions (FAJs)” between cell neighbors. These FAJs contain zyxin, which is

only recruited to cell borders in response to increased cellular tension. Furthermore,

Ena/VASP and testin are only recruited to these structures in cells that have not been

depleted of zyxin. In control cells, the stretch-induced recruitment of zyxin (and thus

Ena/VASP and testin), leads to cytoskeletal strengthening, wherein stress fibers

become thicker and more abundant. When zyxin is depleted, FAJs do not form and the

actin cytoskeleton fails to develop a robust network of stress fibers. Ena/VASP and

15

testin are not completely required for cytoskeletal strengthening to occur, but may make

the network slightly more robust or mediate an indirect interaction between actin, zyxin,

and alpha-catenin. We hypothesize that this in vitro zyxin-dependent actin stabilization

is also at play in vivo during the initiation of apical constriction.

We have previously shown that E cells receive developmental patterning

information from cell intrinsic cues as well as signals from their neighbors. E cells

integrate this information to prompt an increase in tension via MRCK-1 signaling at the

apical surface17. Around this time, these cells also specifically expressing high levels of

zyxin9. We believe that these events must align to properly regulate the timing of apical

constriction initiation during C. elegans gastrulation.

In our model (Figure 9), zyxin transcripts begin accumulating within E cells while

MRCK-1 becomes apically enriched and begins activating myosin. While myosin

activation produces tension within the apical cortex before constriction begins, zyxin is

being translated and folded. The tensile cortex can then recruit folded zyxin to apical

junctions. Once sufficient levels of zyxin have accumulated and integrated themselves

into FAJs along with binding partners Ena/VASP and testin, apical constriction will

begin. We believe this model can help explain the observed increase in slippage rate

after zyxin for depletion in gastrulating embryos.

Future Directions

The following proposed experiments will help address specific questions

regarding zyxin’s role in apical constriction. One key question is whether zyxin directly

physically acts to link the cytoskeleton to adhesive structures at cell junctions and how.

What binding partners are necessary to allow zyxin to sense strain, move to cell

16

contacts, and form connections between components of clutch complex? We also aim

to address the crucial question of how developmental regulation guides the progression

of this event with temporal precision.

In ascertaining the role of zyxin in apical constriction, we will rely on the

CRISPR/Cas system to genetically mutate zyxin, add fluorescent tags, alter cell fate,

and change expression pattern to test our model. By attaching a fluorescent tag to

zyxin, we will be able to measure its levels while also tracking its movement throughout

the embryo. Our expectation is that the timing of zyxin accumulation and junctional

recruitment in E cells will be coordinated with the onset of apical constriction. Below we

list several experiments that will test the hypotheses stemming from our model (Figure

9).

Genetic truncations and point mutations will reveal which domains are crucial to

zyxin’s function. Zyxin’s N-terminus has been shown to interact with actin, while its C-

terminal LIM domains bind with partners like Ena/VASP and testin25. An N-terminal

truncation should keep zyxin from interacting with tensile actin, thus preventing it from

sensing stretch or strengthening the apical cortical meshwork. A C-terminal truncation

would prevent binding with other cytoskeletal interactors, but likely has only a marginal

effect on cytoskeletal strengthening. If these hypotheses hold true, embryos with zyxin

N-terminal truncations will experience abnormally high myosin/membrane slippage

during late stages of apical constriction, while embryos with the C-terminal truncation

will look relatively normal.

A crucial component of our model is that zyxin’s action, under strict

developmental regulation, acts with precise timing. We propose that this is due to the

17

timing of its expression. To test this hypothesis, we will genetically encode zyxin’s

expression to be under control of the med-1 promoter), resulting in premature zyxin

expression. In this scenario, zyxin will accumulate in E cells several minutes before

MRCK-1 is able to activate myosin at the apical surface. Once myosin activation occurs,

however, we would expect the increase in apical tension to result in near-immediate

recruitment of zyxin to cell junctions. In short, apical constriction would initiate several

minutes early. These experiments will be crucial to explain how developmental

regulation instructs the precise timing of gastrulation in C. elegans.

Our model proposes a system in which we are able to comprehensively establish

a connection between developmental patterning and the physical mechanics of early

tissue morphogenesis. Forming this connection requires a set of experiments that will

alter cell fates within the developing embryo. These experiments will utilize the CRISPR

system to make MS cells express the E cell fate, for example. We hypothesize that

these “ectopic E cells” will express zyxin around the time of gastrulation while also

experiencing increased surface tension, ultimately undergoing apical constriction. These

data would provide support to the idea that zyxin and its developmental regulation are

the keystone to the initiation of apical constriction. It is alternatively possible that end-3

expression is absolutely required to link spatial information and force generation and

inform cells to constrict. If this is true, early expression of zyxin would not lead to

premature apical constriction.

If the proposed experiments are unable to substantiate zyxin’s participation in

apical constriction, we still have many promising candidates from our screen to test.

Alpha-catulin (CTN-1) has domains resembling parts of both alpha-catenin and vinculin.

18

Since alpha-catenin is an integral part of CCCs30 and vinculin has been implicated in

FAJ formation33, this is an intriguing candidate for future study. HUM-8 is an

unconventional myosin that resembles human Myosin 6. This motor protein is involved

in epithelial morphogenesis via its interaction with vinculin at cell borders and allows

cells to form cohesive contacts39. A formin-related protein, CYK-1, can assemble actin

filaments34 and is directly involved in cytokinesis35,36. This direct mechanical action on

the cytoskeleton makes it an attractive candidate. For many of our candidates, we only

can distinguish a few protein domains. Proteins with kinase or phosphatase domains

seem important, though their involvement in apical constriction, if any, is likely to be

more indirect.

There is no doubt that a number of experiments are still necessary to fully

demonstrate how tensile cortical actin can mechanically effect a cell shape change with

precise developmental timing in a living organism. Our data have begun to unravel the

linkage between development and biomechanics at the cellular level in vivo. With

single-cell transcriptomic data informing a genetic screen, and with the ability to watch

molecular dynamics in vivo, we have provided evidence suggesting an important role for

zyxin in apical constriction. We believe the experiments proposed above will further

illuminate its role in connecting contractile actomyosin to adhesive structures at cell

junctions. Zyxin likely acts with one or more partners throughout this process, and the

results from our genetic screen will be a valuable source of information when

deconstructing partial redundancies within the system. These data provide us with the

beginnings to a course of research that will help us form a thorough understanding of

the overlap between mechano- and developmental biology.

19

APPENDIX 1: TABLE 1 – RESULTS OF RNAi SCREEN RESULTS

KD target

Stage

(cells)

Enrichment

(Mean

RPKM in E

over MS

Proposed

Function29

Genetic

Background

MS

div

E

intern

(min)

MS

div

E

div

(min)

% Gad

(n/n)

Negative

Control

-- -- -- N2 (wild-

type)

16.2

2.0

19.9

1.7

0

(20/20)

acp-2 16 878.64/0.2 acid phosphatase N2 (wild-

type)

18.8

3.4

19.3

1.6

12.5

(2/16)

add-1 8 8.2/0.0 alpha-adducin:

cytoskeleton

signaling

N2 (wild-

type)

15.7

2.2

20.5

3.8

7.7

(1/13)

C46E10.8 8 108.94/0.0 zinc-finger protein N2 (wild-

type)

16.5

5.7

22.8

4.8

50.0

(6/12)

C29F7.2 16 955.98/0.45 kinase-like domains N2 (wild-

type)

19.1

4.8

23.5

4.8

27.3

(3/11)

C26F1.1 16 699.64/0.57 novel N2 (wild-

type)

17.0

7.7

18.7

4.8

0 (0/9)

ctn-1 8 48.48/0.05 alpha-catulin:

cytoskeleton,

adhesions

N2 (wild-

type)

20.0

8.4

27.6

7.7

20.0

(2/10)

cyk-1 8 1.08/12.3 formin, actin

polymerization

N2 (wild-

type)

18.3

7.9

18.8

4.5

23.1

(3/13)

dve-1 16 857.9/0.0 CMP domain, RAS

signaling

N2 (wild-

type)

13.8

1.7

18.0

2.6

0 (0/4)

F25D7.5 8 40.93/0.07 C-type lectin fold N2 (wild-

type)

14.6

2.9

19.2

0.8

0 (0/5)

F49E10.4 8 134.5/0.42 mitochondria-eating

protein

N2 (wild-

type)

20.5

4.5

26.4

3.6

0 (0/13)

grdn-1 8 9.8/0.0 actin-associated N2 (wild-

type)

11.6

10.8

19.6

7.0

33.3

(3/9)

H24G06.1 8 45.86/0.0 mitogen-activated

protein kinase

binding protein

N2 (wild-

type)

15.0

3.6

20.3

.7

20.0

(2/5)

hum-8 16 26.48/0.0 unconventional

myosin

N2 (wild-

type)

24.8

7.9

20.6

1.7

66.7

(6/9)

let-4 8 25.06/0.0 organizes ECM N2 (wild-

type)

11.9

2.3

19.6

4.0

0 (0/17)

pssy-1 8 137.97/1.94 phosphatidylserine

synthase

N2 (wild-

type)

17.0

5.2

23.6

2.4

7.1

(1/14)

R05D3.2 16 32.49/0.2 LMBR1 human

ortholog – Shh

signaling

N2 (wild-

type)

23.4

7.7

23.5

2.9

51.3

(7/13)

20

KD target

Stage

(cells)

Enrichment

(Mean

RPKM E /

MS)

Proposed

Function29

Genetic

Background

MS

div

E

intern

(min)

MS

div

E

div

(min)

% Gad

(n/n)

R06B10.2 8 6.19/.16 tyrosine

phosphatase

N2 (wild-

type)

23.7

1.2

22.6

5.6

71.4

(5/7)

T14E8.1 8 13.7/0.0 tyrosine kinase N2 (wild-

type)

19.4

5.3

21.3

1.6

18.2

(2/12)

tnc-2 8 54.46/0.32 troponoin N2 (wild-

type)

16.3

2.9

19.3

1.0

0 (0/9)

Y53C10A.

10

8 19.16/0.0 novel N2 (wild-

type)

17.5

3.1

22.4

2.1

8.3

(2/12)

Y57G11C.

6

8 32.48/0.18 tyrosine

phosphatase

N2 (wild-

type)

18.2

2.6

23.1

1.1

30

(3/10)

zig-5 8 27.05/0.0 secreted

immunoglobulin

N2 (wild-

type)

17.2

4.3

23.0

4.1

37.5

(3/8)

zyx-1 8 75.28/0.21 zyxin, focal

adhesions

N2 (wild-

type)

18.6

4.0

22.8

5.4

25

(5/20)

Negative

Control –

sax-7

(eq1)

-- -- -- sax-7 (eq1) 16.7

2.3

21.3

1.8

0

(0/7)

btb-15 16 132.4/1.9 zinc-finger protein sax-7 (eq1) 16.0

1.4

19.5

1.3

0

C29F7.2 16 955.98/0.45 kinase-like domains sax-7 (eq1) 18.7

18.8

21.6

6.7

28.6

(2/7)

C26F1.1 16 699.64/0.57 novel sax-7 (eq1) 29.7

12.0

23.8 66.7

(4/6)

cyk-1 8 1.08/12.3 formin, actin

polymerization

sax-7 (eq1) 23.0

12.8

24.3

5.6

73.3

(11/15)

acp-6 16 72.27/0.0 acid phosphatase sax-7 (eq1) 18.0

4.1

22.2

1.6

20.0

(1/5)

F49E10.4 8 134.5/0.42 mitochondria-eating

protein

sax-7 (eq1) 234 317 9.1

(1/11)

grdn-1 8 9.8/0.0 actin-associated sax-7 (eq1) 26.5

16.6

24.3

5.6

56.5

(13/23)

hex-5 16 220.17/0.0 hexosaminidase sax-7 (eq1) 23.9

3.4

27.0

6.8

33.3

(3/9)

kin-9 8 45.82/0.12 tyrosine kinase sax-7 (eq1) 18.7

1.2

22.3

3.7

50.0

(2/4)

let-4 8 25.06/0.0 organizes ECM sax-7 (eq1) 12.4

1.5

18.0

1.0

0 (0/4)

tnc-2 8 54.46/0.32 troponin sax-7 (eq1) 20.5

3.5

23.7

1.2

33.3

(1/3)

zyx-1 8 75.28/0.21 zyxin, focal

adhesions

sax-7 (eq1) 24.4

11.2

24.5

6.2

66.7

(10/15)

21

KD target

Stage

(cells)

Enrichment

(Mean

RPKM E /

MS)

Proposed

Function29

Genetic

Background

MS

div

E

intern

(min)

MS

div

E

div

(min)

% Gad

(n/n)

ZK1053.3 8 53.07/0.0 novel sax-7 (eq1) 23.7

11.6

24.3

1.2

33.3

(1/3)

Negative

Control –

ced-5

(n1812)

-- -- -- ced-5

(n1812)

17.4

1.9

20.0

2.2

0

(0/5)

acp-2 8 878.64/0.2 acid phosphatase ced-5

(n1812)

33.7

9.8

25.8

7.2

69.2

(9/13)

btb-15 16 132.4/1.9 zinc-finger protein ced-5

(n1812)

22.6

3.3

19.5

2.1

40.0

(2/5)

F49E10.4 8 134.5/0.42 mitochondria-eating

protein

ced-5

(n1812)

23.7

6.4

19.2

3.3

83.3

(5/6)

grdn-1 8 9.8/0.0 actin-associated ced-5

(n1812)

16.8

5.7

22.0

5.5

0 (0/4)

tnc-2 8 54.46/0.32 troponin ced-5

(n1812)

32.4

8.5

22.9

5.8

72.7

(8/11)

zyx-1 8 75.28/0.21 zyxin, focal

adhesions

ced-5

(n1812)

27.6

9.2

22.2

3.0

62.5

(5/8)

Table 1 – Results of RNAi screen. “KD target” column gives gene name or Cosmid ID

for each targeted gene. “Enrichment” and “Stage” columns show data from Tintori et al9,

comparing Reads Per Kilobase per Million (RPKM) between an E cell precursor/MS cell

precursor pair (8 cell stage of embryogenesis) or an E cell/MS cell pair (16 cell stage).

Three genetic backgrounds were used to avoid potential redundancies in the system.

Sax-7 (eq1) is a null allele of a Cell Adhesion Molecule and Ced-5 (n1812) is a null

allele of a cell engulfment protein involved in Ras signaling. The column “MS division

E internalization” shows the time it takes an E cell to fully internalize (see text for

definition) following MS division. +/- represents 95% confidence interval. The column

“MS division E division” displays how long after MS division E cells divide with +/-

representing 95% confidence. The “% Gad” column shows the proportion of dsRNA-

treated embryos that were Gastrulation defective.

22

APPENDIX 2: FIGURES

Figure 1. Apical constriction is a conserved process in C. elegans gastrulation as well as vertebrate neural tube formation. The cells highlighted in green are E cells in C. elegans and neural plate cells in vertebrates and undergo apical constriction. Myosin is enriched and activated (red) in the apical cortex of specific cells, eliciting tension3,8 Adapted from Bob Goldstein.

23

Figure 2. C. elegans embryos achieve gastrulation via apical constriction. (A) Apical constriction initiates when the tensile actomyosin network physically connects to apical adherens junctions. This connection appears to form with precise timing in gastrulating C. elegans embryos. (B) shows a ventral view of a gastrulation-stage embryo, with E cells (Ea, anterior; Ep, posterior) pseudocolored in green undergoing apical constriction. Adapted from Martin & Goldstein 2014 and Lee et al. 2006.

24

Figure 3. Pseudo-kymograph of apical myosin (green) and membrane (red) dynamics during apical constriction. (A) Apical membrane trace of an E cell in the “Early” Stage at 0 and +3 min. (A’) Apical membrane trace of an apically constricting E cell in the “Late” stage at 0 and +3 min. (B), (B’) show the molecular dynamics associated with (A) and (A’) respectively. Early (A, B), there is a large degree of slippage between myosin (green) and membrane (red). Late (A’, B’), there is a low degree of slippage, and constriction proceeds. Adapted from Roh-Johnson et al. 2012

25

Figure 4: The temporally-regulated link between adherens junctions and F-actin. The molecular link between the cadherin-catenin complex and F-actin in apically constricting cells remains a mystery, and continues to be a source of controversy. From Bob Goldstein

26

Figure 5. Choosing candidates based on expression profile in the early embryo. (A) Genes with transcriptional enrichment in a given cell (8-cell stage E cell progenitor here) are colored red. Candidates considered were chosen from within the circle. (B) Zyxin is highly enriched in the E cell progenitor cell at 80 RPKM at the 8-cell stage, compared to its low levels in all other neighboring cells. Adapted from Tintori et al. (2016)9.

27

Figure 6. Results of RNAi screen. Graphical Representation of data from Table 1. Refer to Table 1 legend for details. (A) Timing of E cell Internalization, corresponding to the “MS div > E intern (min)” Column. Red Bars represent a gastrulation defect incidence in >25% of tested embryos. (B) Timing of E cell division, corresponding to the “ MS div > E div (min)” Column.

28

Figure 7. E cells divide at ventral surface in Gad embryos. Lateral mounted embryos at the gastrulation stage. (A, A’) Wild-type progression of gastrulation where E cells (top, red outline) divide after fully internalizing. (B, B’) A Gad embryo after targeting zyxin for depletion, where an E cell daughter is born at the ventral surface. (A, B) 90 minutes after fertilization. (A’, B’) 120 minutes after fertilization.

29

Figure 8. Slippage Rate remains abnormally high after zyxin depletion. (A,B) Snapshot of ventrally mounted wild-type embryos in Early and Late stage, respectively. Yellow arrows represent myosin puncta movements in E cells. Red arrows represent movement of associated membrane. (C) Myosin’s rate of centripetal flow does not change from Early to Late stage. The dsRNA treatment targeting zyxin does also not cause a change in myosin flow rate. (D) In both wild-type (blue circiles) and zyxin-depleted (red squares) embryos, there is a high degree of slippage between myosin and membrane in Early Stage E cells. The slippage rate in Late Stage E cells zyxin-depleted embryos remains high abnormally. (E) The same embryos analyzed by kymography in C and D were independently measured using the ImageJ mTrackJ plugin in a double-blind study (credit: Terrance Wong). (C,D,E) n = 5 wild-type and 6 zyx-1-targeted embryos.

30

Figure 9. Our model of zyxin’s action during apical constriction. In our model, an early increase in tension fails initiate constriction because the actomyosin network is not connected to the cadherin-catenin complex. Once zyxin has accumulated to sufficient levels, it forms a complex between actin, Ena/VASP (UNC-34), Testin (TES-1), and the CCC at apical cell junctions, allowing constriction to proceed.

31

REFERENCES

1. Keller R, Davidson LA, Shook DR. How we are shaped: the biomechanics of Gastrulation. Differentiation. 2003;71(3):171-205.

2. Lecuit T, Lenne PF, Munro E. Force generation, transmission, and integration during cell and tissue morphogenesis. Annu Rev Cell Dev Biol. 2011;27:157-84

3. Martin AC, Goldstein B. Apical constriction: themes and variations on a cellular mechanism driving morphogenesis. Development. 2014 May; 141(10): 1987–1998.

4. Sawyer JM, Harrell JR, Shemer G, Sullivan-Brown J, Roh-Johnson M, Goldstein B. Apical Constriction: a cell shape change that can drive morphogenesis Dev Biol. 2010;341(1):5-19.

5. Roh-Johnson M, Shemer G, Higgins CD, McClellan JH, Werts AD, Tulu US, Gao L, Betzig E,

Kiehart DP, Goldstein B. Triggering a Cell Shape Change by Exploiting Preexisting Actomyosin Contractions Science. 2012;335(6073):1232-5.

6. Lee JY, Marston DJ, Walston T, Hardin J, Halberstadt A, Goldstein B. Wnt/Frizzled Signaling Controls C. elegans Gastrulation by Activating Actomyosin Contractility. Curr Biol. 2006 Oct 24; 16(20): 1986–1997.

7. Lee JY, Goldstein B. Mechanisms of cell positioning during C. elegans gastrulation.

Development. 2003 Jan; 130(2): 307–320. 8. Sullivan-Brown JL, Tandon PD, Bird KE, Dickinson DJ, Tintori SC, Heppert JK, Trogden KP,

Orlowski SK, Conlon FL, Goldstein B. Identifying regulators of morphogenesis common to vertebrate neural tube closure and Caenorhabditis elegans gastrulation. Genetics. 2016;202(1):123-39.

9. Tintori SC, Nishimura EO, Golden P, Lieb JD, Goldstein B. A Transcriptional Lineage of the Early

C. elegans Embryo. Dev Cell. 2016 Aug 22; 38(4): 430–444. 10. Yoshigi M, Hoffman LM, Jensen CC, Yost HJ, Beckerle MC. Mechanical Force mobilizes zyxin

from focal adhesions to actin filaments and regulates cytoskeletal reinforcement. J Cell Biol. 2005;171(2):209-15.

11. Smith MA, Blankman A, Deakin NO, Hoffman LM, Jense CC, Turner CE, Beckerle MC. LIM

domains target actin regulators paxilin and zyxin to sites of stress fiber strain. PLoS One. 2013;8(8):e69378

12. Hoffman LM, Jensen CC, Chaturvedi A, Yoshigi M, Beckerle MC. Stretch-induced actin

remodeling requires targeting of zyxin to stress fibers and recruitment of actin regulators. Mol Biol Cell. 2012;(10):1846-59.

13. Swaminathan V, Waterman CM. The molecular clutch model for mechanotransduction evolves.

Nat Cell Biol 2016 Apr 27;18(5):459-61.

14. Botto LD, May K, Fernhoff PM, Correa A, Coleman K, Rasmussen SA, Merritt RK, O'Leary LA, Wong LY, Elixson EM, Mahle WT, Campbell RM. Trends of selected malformations in realtion to folic acid recommendations and fortification: an international assessment. Pediatrics. 2003;112:101-7.

15. Dickinson DJ, Ward JD, Reiner DJ, Goldstein B. CRISPR-based methods for Caenorhabditis elegans genome engineering. Nat Methods. 2013;10(10):1028-34

32

16. Dickinson DJ, Pani AM, Heppert JK, Higgins CD, Goldstein B. Stream-lined genome engineering

with a self-excising drug selection cassette. Genetics. 2015;200(4):1035-49 17. Marston DJ, Higgins CD, Peters KA, Cupp TD, Dickinson DJ, Pani AM, Moore RP, Cox AH,

Kiehart DP, Goldstein B. MRCK-1 Drives Apical Constriction in C. elegans by Linking Developmental

18. Patterning to Force Generation. Curr Biol. 2016 Aug 22;26(16):2079-89. 19. Smith MA, Hoffman LM, Beckerle MC. LIM proteins in actin cytoskeleton mechanoresponse.

Trends Cell Biol. 2014;24(10):575-83. 20. Martin AC, Kaschube M, Wieschaus EF. Pulsed actin-myosin network contractions drive apical

constriction. Nature. 2009 Jan 22; 457(7228): 495. 21. Dawes-Hoang RE, Parmar KM, Christiansen AE, Phelps CB, Brand AH, Wieschaus EF. Folded

gastrulation, cell shape change, and the control of myosin localization. Development. 22. 2005;132(18):4165-78. 23. Sawyer JM, Glass S, Li T, et al. Overcoming Redundancy: An RNAi Enhancer Screen for

Morphogenesis Genes in Caenorhabditis elegans. Genetics. 2011;188(3):549-564. doi:10.1534/genetics.111.129486.

24. Nance J, Munro EM, Priess JR. Cortical flows powered by asymmetrical contraction transport

PAR proteins to establish and maintain anterior-posterior polarity in the early C. elegans embryo. Dev Cell. 2004;7(3):413-24.

25. Nishimura T, Takeichi M. Shroom3-mediated recruitment of Rho kinases to the apical cell

junctions regulates epithelial and neuroepithelial planar remodeling. Development. 2008;135(8):1493-502.

26. Oldenburg J, van der Krogt G, Twiss F, Bongaarts A, Habani Y, Slotman JA, Houtsmuller

A, Huveneers S, de Rooij J. VASP, zyxin and TES are tension-dependent members of Focal Adherens Junctions independent of the α-catenin-vinculin module. Sci Rep. 2015;5:17225.

27. Stachiowak MR, Smith MA, Blankman E, Chapin LM, Balcioglu HE, Wang S, Beckerle MC,

O’Shaughnessey B. A mechanical-biochemical feedback loop regulates remodeling of the actin cytoskeleton. Proc Nat. Acad. 2014:111(49):17528-33

28. Martin AC, Gelbart M, Fernandez-Gonzalez R, Kaschube M, Wieschaus EF. Integration of

contractile forces during tissue invagination. J Cell Biol. 2010. 188(5):735-49 29. Kamath RS, Martinez-Campos M, Zipperlen P, Fraser AG, Ahringer J. Effectiveness of specific

RNA-mediated interference through ingested double-stranded RNA in Caenorhabditis elegans. Genome Biol. 2001. 2(1);2001

30. S. Brenner, The genetics of Caenorhabditis elegans. Genetics 77, 71 (1974)

31. WormBase web site, http://www.wormbase.org, version WS255, August 1, 2016. 32. Niessen CM, Leckband D, Yap AS. Tissue organization by cadherin adhesion molecules:

dynamic molecular and cellular mechanisms of morphogenetic regulation. Physiological reviews. 2011;91(2):691-731.

33

33. Abu Taha A, Schnittler H-J. Dynamics between actin and the VE-cadherin/catenin complex: Novel aspects of the ARP2/3 complex in regulation of endothelial junctions. Cell Adhesion & Migration. 2014;8(2):125-135.

34. Craig D. Buckley, Jiongyi Tan, Karen L. Anderson, Dorit Hanein, Niels Volkmann, William I. Weis,

W. James Nelson, Alexander R. Dunn. The minimal cadherin-catenin complex binds to actin filaments under force. Science. 2014 Oct 31; 346(6209): 1254211

35. Huveneers S, Oldenburg J, Spanjaard E, et al. Vinculin associates with endothelial VE-cadherin

junctions to control force-dependent remodeling. The Journal of Cell Biology. 2012;196(5):641-652. doi:10.1083/jcb.201108120.

36. Neidt EM, Skau CT, Kovar DR. The Cytokinesis Formins from the Nematode Worm and Fission Yeast Differentially Mediate Actin Filament Assembly. The Journal of Biological Chemistry. 2008;283(35):23872-23883.

37. Gönczy P, Schnabel H, Kaletta T, Amores AD, Hyman T, Schnabel R. Dissection of Cell Division Processes in the One Cell Stage Caenorhabditis elegans Embryo by Mutational Analysis . The Journal of Cell Biology. 1999;144(5):927-946.

38. Jordan SN, Davies T, Zhuravlev Y, Dumont J, Shirasu-Hiza M, Canman JC. Cortical PAR polarity

proteins promote robust cytokinesis during asymmetric cell division. The Journal of Cell Biology. 2016;212(1):39-49. doi:10.1083/jcb.201510063.

39. Grana TM, Cox EA, Lynch AM, Hardin J. SAX-7/L1CAM and HMR-1/cadherin function

redundantly in blastomere compaction and non-muscle myosin accumulation during C. elegans gastrulation. Developmental biology. 2010;344(2):731-744.

40. Wu YC, Horvitz HR. C. elegans phagocytosis and cell-migration protein CED-5 is similar to

human DOCK180. Nature. 1998 392: 501–504 41. Madhavi P. Maddugoda, Matthew S. Crampton, Annette M. Shewan, Alpha S. Yap. The Journal

of Cell Biology Jul 2007, 178 (3) 529-540.